Douglas Hagedorn and Nicole Timoshenko
 3 - Research Approach: Objective Three


Develop a GIS-based model that incorporates a Weighted Overlay process to assess land suitability and MCE process to target and prioritize locations suitable for Alvar conservation initiatives.


The purposed of this study is to justify the scaling of each of the factors identified in objective one and two and to justify the
criteria weighting determined between factors.  Weighting criteria allows for preference or priority to be assigned to specific criteria
for MCE (Heywood et al., 2002).  Weighting are chosen according to the relative importance of each criteria.  A MCE model combines criteria maps according to these user-defined weightings to produce suitability maps (Heywood et al., 2002).  Outlined below is a list of criteria to be used in objective three, as well as justification of these variables and their relative importance.  Table 3-1 provides is a concise reference table of the information discussed below.


Stage One: Identifying Land Suitable for Conservation

A MCE model uses the index values which have been calculated from a composite map to produce a map with ranked areas.  Table 3-1 lists the variables under consideration, as well as the maximum distance contributing to anthropogenic risk, and the standardization assigned to each factor.

1) Distance from transportation infrastructure:
As was discussed in objective one, the size of the road will determine the extent of anthropogenic disturbance and relative risk (Trombulak and Frissel, 2000) a result, highway, collector and local roads will have different distance criteria. 
This study will standardize straight line distances from major roads on a scale of 0-100, where higher values denote areas which are closer to major roads, and thus face greater anthropogenic risks.  The straight line distance will be evaluated only to a distance of 1500 m, since the greatest risk of sediment or pollutant transport is just beyond 1000 m (Forman and Alexander, 1998).
The anthropogenic risk posed by an intermediate road is less than a major road (Trombulak and Frissel, 2000) and thus straight line distance will be evaluated only to a distance of 1000 m for intermediate roads.  The distances will be standardized on a scale of 0-100, where higher values denote areas which are closer to intermediate roads, and thus face greater anthropogenic risk.
The same principle will be applied to minor roads.  Straight line distances will be evaluated only to a distance of 600 m, and standardized on a scale of 0-100, where higher values denote areas which are closer to minor roads.

2) Distance from agricultural land
Pesticide or herbicide runoff is the major concern for land in proximity to agricultural land (Lovell and Sulivan, 2004).  Straight line distances will be evaluated to a distance of 1000 m, since beyond this the effects of hydrological runoffs are minimal (Forman and Alexander, 1998).  The distances will be standardized on a scale of 1-100, with higher values assigned to land which faces greater anthropogenic risk due to proximity to agriculture land use.

3) Distance from urban centres
Urban centres are a threat to natural systems for a variety of reasons as identified in objective one.  Straight line distance from urban centres will be considered to a distance of 1000 m, and standardized on a scale of 0-100, with higher values assigned to land closer to urban centres.  This distance is consistent with the hydrological runoff effects that could be expected from a city (Forman and Alexander, 1998), and represents the average length of a walk journey and hence represents a realistic distance in which we could expect anthropogenic disturbance in natural areas due to hikers (National Statistics, 1997). 

For the purpose of this study, we will identify functional conservation area land cover.  These areas are defined by their ability to maintain focal ecosystems, species, and support ecological processes in their natural context (Poiani et al., 2000). Objective one identified appropriate land cover as a constraint towards assessing land for conservation purposes.  Within this restriction, areas that are not suited for conservation include tracts of land classified as settlements and developed land, pastures, abandoned fields, and cropland (Benedict and McMahon, 2002).  These land values will be restricted, receiving a value of 0, and all other classes will be considered unrestricted appropriate land cover and receive a value of 1.  This constraint map will determine which tracts of land are deemed suitable for conservation efforts based on the land cover constraints. 
 
In order to assign criteria weights to each factor, the Pairwise Comparison method was utilized.  Factors were considered in pairs and their relative importances were determined.  It was determined that distances from transportation infrastructure, agricultural land use, and urban centers were equally important due to similarities in the distances of their effects.  The only factors which were assigned relative values were the road factors within the transportation network, due to the differences in their effects as outlined in objective one.  Table 3-2 provides the pairwise comparison matrix describing the relative importance of one factor over the others for all possible factor pairs, the individual factor weights and the total weights.



Stage Two:  Targeting and Prioritizing Potential Alvar Conservation Sites

To target and prioritize potential Alvar conservation sites, a layer depicting parcels of conservable land for evaluation, values for the six factors outlined in Objective two that describe attributes of each parcel, and a MCE model with factor weightings must be developed.

The first step involved in the targeting and prioritization of potential Alvar conservation sites involves utilizing a number of analytical processes that will manipulate and build upon the data contained in the Ontario Land Coverage dataset to produce the attributes needed for the MCE model. Figure 3-2 depicts the process layout and flow for this first step of this stage two of the research wherein the evaluation land parcel layer and and attributes values are created and calculated.  This diagram should be read from top to bottom so that the processes listed in the legend are seen to be applied upon the Manitoulin Land coverage data in order to create evaluation parcels (Process 3-6), calculate attributes for these parcels (Processes 3-6 through 3-12) that will ultimately be consolidated in the attribute table of the Final Evaluation Parcel layer.  The application of this framework on the Manitoulin island study site will be described in more detail in Objective Four.   


The second step involved in the targeting and prioritization of potential Alvar conservation sites involves the application of a multiple criteria evaluation procedure in order to provide each targeted evaluation parcels created in the first step with a conservation priority value that can used to create a series of parcels ranked in ascending order.  Figure 3-3 depicts the M.C.E. model and relative factor weights used to perform the final prioritization of the evaluation parcels. This diagram should be read from top to bottom so that the attributes of the Final Evaluation parcel layer shown at the top are used in conjuntion with Rank Sum Criteria weights, which are  calculated in Table 3-3, to create an output map that visually depicts the relative conservation priority value of each parcel using a graduated color scheme.  The application of this model on the Manitoulin Island study site will be described in Process 3-13 in Objective Four.
 
BackHomeForward